阈值化分割算法及其在显著物体检测中的应用研究
时间:2018-09-04 17:40 来源:毕业论文 作者:毕业论文 点击:次
| 摘要： 图像分割是图像识别和图像分析的基本前提步骤，图像分割的质量好坏直接影响后续图像处理的效果，甚至决定成败。因此，图像分割在数字图像处理技术中占有非常重要的地位。物体检测也是图像处理领域的研究热点之一，它有极广泛的应用，例如飞机航拍或卫星图像中的道路检测，监控视频中的车辆行人检测，CT图像中的肿瘤检测等等。本课题在对数字图像分割研究探讨的基础上，研究适合显著物体检测的阈值分割算法。重点在于将经典阈值分割方法运用于显著物体检测，并通过MATLAB编程调试实现，结合算法复杂度以及实验结果，对算法进行性能分析。总的来说，阈值分割的方法很多，每一种方法几乎都有其独特的优点和实际应用的背景。实际应用中，阈值分割需要和其他方法相互结合使用，才能获得最佳或满意的分割结果。27725 毕业论文关键词： 阈值分割；物体检测；最大类间方差法；矩量保持法；算法效率 Thresholding Segmentation Algorithm and its application on Visual Object Detection Abstract: Image Segmentation is the basic premise of Image Recognition and Image Analysis. The quality of Image Segmentation affects the effect of Image Processing directly, even determine the success or failure. Therefore, Image Segmentation plays a very important role in Digital Image Processing Technology. Object Detection is one of the focus areas of the Image Processing. It has a very wide range of applications, like aircraft aerial or satellite images road detection, surveillance video of the vehicle pedestrian detection, CT images of tumor detection, etc. This subject is on the basis of the research of Digital Image Segmentation, research a suitable Thresholding Segmentation Algorithm of Significant Object Detection. Focuses on classical thresholding method is applied to detect significant objects, and the realization by MATLAB programming and debugging, combined with the complexity of the algorithm and experimental results, analysis algorithm performance. In short, there are many threshold segmentation method, each method has its own unique background almost advantages and practical applications. In practical applications, thresholding and other methods need to be used in conjunction with each other to get the best or satisfactory segmentation results. Keywords: thresholding segmentation; object detection; Otsu; moment-preserving; efficiency of algorithm 目录 摘要 i Abstract ii 目录 iii 1 绪论 1 1.1 研究背景与意义 1 1.2 国内外研究现状与水平 2 1.2.1 图像阈值分割的技术发展 2 1.2.2 图像阈值分割的研究现状与发展趋势 4 1.2.3 物体检测的技术发展和研究现状 4 1.3 本文的组织结构 5 2 阈值分割算法 6 2.1 Otsu最大类间方差法 7 2.2 Tsai矩量保持法 8 3 系统设计 10 3.1 运行环境 10 3.2 界面设计 11 4 结果分析 21 5 总结 23 5.1 工作总结 23 5.2 存在的问题及展望 23 致谢 24 参考文献 25 1 绪论 (责任编辑：qin) |